Day-ahead Electricity Price forecasting using Wavelets and Weighted Nearest Neighborhood

Price forecasting has been at the center of intense studies since the introduction of competition in electricity industry. Price forecasts are a fundamental input to an energy company's decision making and strategy development. The present approach is an attempt to forecast day-ahead electricity prices using time series of historical data. A combination of weighted nearest neighborhood and wavelets is used to forecast the next day electricity prices. The methodology is applied to historical data pertaining to California electricity market. The performance of the method is discussed with mean absolute percentage error (MAPE).